Major changes: Fixed FlankerRep and TargetRep to include duplications (e.g., 130 followed by 130). Added model with TargetRep included as covariate.

1. FSW

“Current” trial will be included if:
1) Current trial is not rejected due to artifacts.
2) Current trial is not the first trial of the block (i.e. trial 1, 101, 201, 301 etc. are not included).
3) Both current trial and previous trial are correct.

The following electrodes are included: F3, Fz, F4, FC3, FCz, FC4, C3, Cz, C4.

The following subjects were excluded:
- 92 (wrong number of trials, possibly data collection error)
- 64 (<50% accuracy rate)
- 65 (only 187 usable trials (23.3%))

One or more electrodes were bad for subjects 19, 74, and 137 (data from those electrodes is discarded).

Besides that, ERP data is completely missing for subs 36, 37, 39, 40, 42, 43, 46, 51, 53, 54, 56, 57, 78.

Total sample is 36 subjects.

Trigger Codes: Comp: Target Left, Flanker Left = 110.
Comp: Target Right, Flanker Right = 130.
Incomp: Target Left, Flanker Right = 150.
Incomp: Target Right, Flanker Left = 170.

CompleteRep (24.3%) = repetition of same trial (e.g., excludes 110 following 110)
AnyRep (74.8%) = either target or flanker is repeated (only includes 110/130 pair, 150/170 pair)
TargetRep (~50%) = target is repeated (excludes 150/110 pair, 170/130 pair)
FlankerRep (~50%) = flanker is repeated (excludes 150/130 pair, 170/110 pair)

A. TargetRep trials taken out

Plot over whole task

Slopes and estimates of lines are from the MLM, not fitted with OLS.

Simple slopes

Trial is scaled to range from 0 to 8 (instead of 1 to 800) so that the betas associated with trial are a little bigger (but significance testing is unaffected by linear scaling, so the test statistics and p values will be the same as if we used the unscaled Trial variable).

Estimate SE ci95_lower ci95_upper Current Previous Color
0.256 0.0270 0.202 0.310 Compat Compat blue dashed
0.253 0.0285 0.196 0.310 Compat Incompat red dashed
0.448 0.0289 0.391 0.506 Incompat Compat blue solid
0.283 0.0298 0.223 0.342 Incompat Incompat red solid
Fixed effects of Trial and interactions with Trial:

Trial is centered and scaled to range from -4 to 4.

Estimate Std. Error df t value Pr(>|t|)
Trial.begin 0.310 0.014 100124.9 21.695 0.000
Current.e:Trial.begin 0.055 0.014 100142.5 3.879 0.000
Previous.e:Trial.begin -0.042 0.014 100111.0 -2.953 0.003
Current.e:Previous.e:Trial.begin -0.041 0.014 100078.9 -2.848 0.004
Fixed effects at beginning of task:
Estimate Std. Error df t value Pr(>|t|)
(Intercept) -3.292 0.347 35.822 -9.498 0.000
Current.e -0.466 0.113 58.476 -4.135 0.000
Previous.e 0.116 0.101 69.468 1.153 0.253
Current.e:Previous.e 0.098 0.097 74.109 1.009 0.316

Marginal means:
Error bars represent standard error.

Fixed effects at middle of task:
Estimate Std. Error df t value Pr(>|t|)
(Intercept) -2.052 0.342 34.003 -5.998 0.000
Current.e -0.245 0.098 33.521 -2.492 0.018
Previous.e -0.053 0.084 33.630 -0.626 0.536
Current.e:Previous.e -0.065 0.080 33.689 -0.807 0.425

Marginal means:
Error bars represent standard error.

Fixed effects at end of task:
Estimate Std. Error df t value Pr(>|t|)
(Intercept) -0.812 0.347 36.034 -2.339 0.025
Current.e -0.023 0.114 61.761 -0.200 0.842
Previous.e -0.221 0.103 74.320 -2.160 0.034
Current.e:Previous.e -0.227 0.099 79.631 -2.292 0.025

Marginal means:
Error bars represent standard error.

B. CompleteRep trials taken out

Plot over whole task

Slopes and estimates of lines are from the MLM, not fitted with OLS.

Simple slopes

Trial is scaled to range from 0 to 8 (instead of 1 to 800) so that the betas associated with trial are a little bigger (but significance testing is unaffected by linear scaling, so the test statistics and p values will be the same as if we used the unscaled Trial variable).

Estimate SE ci95_lower ci95_upper Current Previous Color
0.258 0.0270 0.204 0.312 Compat Compat blue dashed
0.268 0.0202 0.228 0.309 Compat Incompat red dashed
0.450 0.0203 0.410 0.491 Incompat Compat blue solid
0.289 0.0297 0.229 0.348 Incompat Incompat red solid
Fixed effects of Trial and interactions with Trial:

Trial is centered and scaled to range from -4 to 4.

Estimate Std. Error df t value Pr(>|t|)
Trial.begin 0.315 0.012 150331.4 25.563 0.000
Current.e:Trial.begin 0.052 0.012 150158.5 4.207 0.000
Previous.e:Trial.begin -0.039 0.012 150122.3 -3.149 0.002
Current.e:Previous.e:Trial.begin -0.046 0.012 150304.4 -3.742 0.000
Fixed effects at beginning of task:
Estimate Std. Error df t value Pr(>|t|)
(Intercept) -3.346 0.357 35.268 -9.379 0.000
Current.e -0.441 0.108 52.508 -4.072 0.000
Previous.e 0.093 0.081 79.979 1.148 0.254
Current.e:Previous.e 0.154 0.094 61.444 1.644 0.105

Marginal means:
Error bars represent standard error.

Fixed effects at middle of task:
Estimate Std. Error df t value Pr(>|t|)
(Intercept) -2.084 0.353 33.994 -5.895 0.000
Current.e -0.233 0.097 33.888 -2.402 0.022
Previous.e -0.063 0.065 33.441 -0.966 0.341
Current.e:Previous.e -0.030 0.081 33.567 -0.375 0.710

Marginal means:
Error bars represent standard error.

Fixed effects at end of task:
Estimate Std. Error df t value Pr(>|t|)
(Intercept) -0.822 0.357 35.411 -2.301 0.027
Current.e -0.025 0.109 54.923 -0.231 0.818
Previous.e -0.218 0.082 86.475 -2.649 0.010
Current.e:Previous.e -0.215 0.095 65.194 -2.254 0.028

Marginal means:
Error bars represent standard error.

C. TargetRep trials retained, included as covariate

Plot over whole task, separately for TargetRep and non-TargetRep

Slopes and estimates of lines are from the MLM, not fitted with OLS.

Beginning of the task

Current and Previous are effect coded, trial ranges from 0 to 8.

Estimate Std. Error df t value Pr(>|t|)
(Intercept) -3.259 0.347 36.16 -9.398 0.000
Current.e -0.455 0.065 199485.86 -7.018 0.000
Previous.e 0.115 0.065 199486.26 1.777 0.076
Current.e:Previous.e 0.088 0.065 199487.53 1.354 0.176
Estimate Std. Error df t value Pr(>|t|)
Trial.begin 0.311 0.014 199488.9 21.728 0.000
Current.e:Trial.begin 0.055 0.014 199485.9 3.850 0.000
Previous.e:Trial.begin -0.035 0.014 199486.1 -2.451 0.014
Current.e:Previous.e:Trial.begin -0.035 0.014 199487.9 -2.433 0.015
Trial.begin:TargetRep 0.029 0.020 199488.3 1.426 0.154
Current.e:Trial.begin:TargetRep -0.007 0.020 199486.6 -0.326 0.744
Previous.e:Trial.begin:TargetRep -0.011 0.020 199487.4 -0.565 0.572
Current.e:Previous.e:Trial.begin:TargetRep -0.002 0.020 199487.3 -0.093 0.926
Estimate Std. Error df t value Pr(>|t|)
TargetRep -0.129 0.092 199487.6 -1.401 0.161
Current.e:TargetRep 0.008 0.092 199486.3 0.087 0.931
Previous.e:TargetRep -0.022 0.092 199487.3 -0.234 0.815
Current.e:Previous.e:TargetRep 0.208 0.092 199487.5 2.270 0.023
Middle of the task

Current and Previous are effect coded, trial ranges from -4 to 4.

Estimate Std. Error df t value Pr(>|t|)
(Intercept) -2.015 0.342 34.317 -5.887 0.000
Current.e -0.234 0.033 199487.799 -7.105 0.000
Previous.e -0.025 0.033 199488.983 -0.761 0.447
Current.e:Previous.e -0.052 0.033 199488.555 -1.563 0.118
Estimate Std. Error df t value Pr(>|t|)
TargetRep -0.129 0.092 199487.6 -1.401 0.161
Current.e:TargetRep 0.008 0.092 199486.3 0.087 0.931
Previous.e:TargetRep -0.022 0.092 199487.3 -0.234 0.815
Current.e:Previous.e:TargetRep 0.208 0.092 199487.5 2.270 0.023
End of the task

Current and Previous are effect coded, trial ranges from -8 to 0.

Estimate Std. Error df t value Pr(>|t|)
(Intercept) -0.771 0.347 36.374 -2.220 0.033
Current.e -0.014 0.067 199488.214 -0.209 0.834
Previous.e -0.165 0.067 199488.603 -2.457 0.014
Current.e:Previous.e -0.191 0.067 199489.838 -2.836 0.005
Estimate Std. Error df t value Pr(>|t|)
TargetRep -0.129 0.092 199487.6 -1.401 0.161
Current.e:TargetRep 0.008 0.092 199486.3 0.087 0.931
Previous.e:TargetRep -0.022 0.092 199487.3 -0.234 0.815
Current.e:Previous.e:TargetRep 0.208 0.092 199487.5 2.270 0.023